By returning to the historical record of how mass atrocities end, this essay examines three crucial narrative frameworks that inform today’s agenda of the “protection of civilians” in conflict. The evidentiary record of actual cases of mass atrocity demonstrates a broad range of forces—local, national, and regional—that contribute to ending atrocities. Based on comparative evidence, the authors provide a counter-narrative to how mass atrocities end and the dominant civilian protection agenda.

By Alex de Waal, Jens Meierhenrich, & Bridget Conley-Zilkic

On October 20, 2011, the battered body of the deposed Libyan leader, Muammar Qaddafi, was paraded through the streets of Sirte and Misrata. Although the details of his final hours are in question, some facts are clear. There were at least two bullet holes in the 69-year-old’s dead body, one to the head and one to the stomach. He was shot after forces loyal to the National Transitional Council beat him, leaving him cut and bruised. It is also possible that some of his wounds were the result of a NATO air attack on his 100-car convoy as it fled Sirte, the site of his final stand. While a complete official version of his death has yet to be pronounced, journalists located several eyewitnesses whose testimony, in addition to the brutal video that surfaced first on Al-Arabiya[1], indicate that he was executed after being taken into custody. Reuters journalist Rania El Gamal quoted a local commander who asserted that while the leaders of the Interim Transitional National Council (NTC) wanted to keep Gaddafi alive, that “over-enthusiastic” young fighters executed him.[2]

And thus ended the 2011 Libyan Civil War, as well as NATO’s mission, “Unified Protector.” It is not shocking that a civil war would conclude with a furious, over-enthusiastic, and young armed force finishing off a leader who had reigned for more than forty years; there are many such examples throughout history and across the globe. But it ought to give pause for reflection on why an international intervention authorized by UN Security Council Resolution 1973 “to protect civilians and civilian populated areas under threat of attack,” used the cover of an humanitarian imperative in order to take sides on an agenda of regime change.


The story of what enabled this chain of events includes three crucial narrative frameworks that inform today’s agenda of “protection of civilians” in conflict. First is a teleological assumption that the occurrence of attacks against civilians will, unless halted or deterred from outside, inevitably escalate towards genocide. Linked to this is the assumption that wherever violence against civilians is deliberate and systematic, the killing of these civilians is the motive of those inflicting the violence. Second is an epistemological assumption that privileges coercive military operations conducted on humanitarian grounds by international forces. Third is an ethical imperative based on the above teleology and epistemology that forecloses the historical and political discussions of how mass atrocities actually end. This essay examines these three frameworks and, based on comparative evidence, provides a counter-narrative to the dominant civilian protection agenda by returning to the historical record of how mass atrocities end.


The Holocaust and Rwanda cast long shadows over research and policy discussions of genocide. In the case of the Holocaust, the objective of the perpetrator regime was the eradication of the entire European Jewish population, in line with its extreme anti-Semitic ideology, either by killing or otherwise ensuring their deaths (through slave labor, starvation and other conditions of life). The killing of the Jews was one of the Nazis’ principal aims, to the extent that the program of killing consumed resources that would otherwise have been available to the war effort, and thereby decreased the chances of survival, or at least the longevity, of the Third Reich.

In Rwanda, the Hutu Power regime was informed by a racist ideology and similarly motivated to annihilate the entire Tutsi populace. To pursue this aim, it also diverted military resources from the fight against a rebel force associated with the ethnic identity of the primary victims, the Tutsi.

Genocide in both the cases of the Holocaust and Rwanda halted only with the overthrow of the perpetrator regime. This was consistent with the fact that, for both regimes, extermination was stated policy and a priority that equaled, or surpassed, regime survival. For both, the violence had an essentialist logic: it was pursued for its own sake.

The dominance of these two cases is problematic in the study of genocide or mass atrocities, and acutely so in the development of policy. Much of the literature on genocide possesses an overdetermined narrative which begins with the origins of genocide in inter-group discrimination and negative stereotyping, with recent examples using the media to propagate such stereotypes, developing through political exclusion to group-targeted violence and ultimately genocide. An essentialist logic of violence is implied: the perpetrators are seen as desiring the destruction of the target group more than anything else. In its simplest form this is seen as a graduated scale of warnings of genocide that corral the full complexity of conflict and inter-ethnic relations into a one-dimensional slippery slope that leads inexorably to genocide, and reduce the varied instrumental political logics of violence to an evil motive alone. These cases model only two possible outcomes: either a completed extermination of the target group or an external military intervention to bring an end to the killing.

Much of the literature on genocide possesses an over-determined narrative that begins with the origins of genocide in inter-group discrimination and negative stereotyping, with recent examples using the media to propagate such stereotypes, developing through political exclusion to group-targeted violence and ultimately genocide.

In the shadow of this historical legacy, policy debates have been reduced to identifying the warning signs and the point at which warnings and intervention should be triggered, along with establishing the standby military force ready to intervene and the legal regimen under which it can do so legitimately. This is the agenda that has informed today’s civilian protection agenda, as witnessed in debates on humanitarian intervention beginning in the 1990s and continuing with the responsibility to protect (R2P, 2001) and the U.S.-focused Genocide Prevention Task Force (GPTF, 2008). The latter two projects, R2P and GPTF, are also marked by an awareness of the limitations of “genocide” as a framework and a consensus that preventive action is preferable to responding to crises after thousands have been killed.

Genocide is, as legally defined in the 1948 Convention on the Prevention and Punishment of the Crime of Genocide, the commission of specified acts with the intent to destroy in whole or in part an ethnic, national, racial or religious group, as such. Research, policy and activism conducted under its banner have struggled with the strictures of this definition: how would one recognize the requisite intent? What groups are excluded from designated victim groups? What “part” of a whole constitutes a sufficiently significant part that its destruction would pose a serious threat to the group? Additionally, researchers and policymakers recognize that many important contemporary cases of systematic abuses against civilians would not fit this definition, however much it might be stretched.

Hence, attempts have been made to introduce new concepts or to add ethical weight and policy obligations to a host of vocabularies: the suite of crimes included in responsibility to protect, or R2P, (genocide, war crimes, crimes against humanity, and ethnic cleansing), former U.S. Ambassador-At-Large for War Crimes, David Scheffer’s “atrocity crimes,”[3] or the increasingly utilized, yet rarely defined, mass atrocities. Each of these vocabulary innovations attempts to capture the ethical significance and call to action embedded in the term “genocide,” but to loosen its rules to include crimes that involve diverse methods of targeting civilians, motivations short of full extermination, and a broader definition of potential victim groups.

Recent initiatives to formulate UN and U.S. policies agree that the priority should be preventive action, and that mass violence against civilians develops incrementally. However, there are two main difficulties in applying these policy frameworks to prevent genocide. First, there is not enough understanding of why, of the many cases that share commonly acknowledged risk factors for genocide or mass violence against civilians, some develop into such violence while others do not. Second, the atrocity and genocide prevention agendas do not offer appreciably new approaches to established response agendas outside of a rationale for armed intervention.[4]

It should not be surprising that the military response components of R2P and GPTF have garnered the most attention. It is arguable that instead of infusing an atrocities-prevention lens into pre-existing development and democratization programs, these and other efforts to promote early action to prevent atrocities or genocide have unleashed a new and ill-defined paradigm for military intervention. Lacking finely-tuned analytical tools, the shift into a discourse of prevention has the effect of transferring the teleological assumptions of genocide to an even broader array of conditions and cases.

The new vocabularies are perhaps of value in providing a less stringent conceptual framework for analysis of diverse cases of mass violence against civilians, but only if they critically engage the assumptions embedded in genocide.

Jens Meierhenrich has proposed an overall framework for studying genocide termination that challenges this teleology. He argues for disaggregating genocidal acts, campaigns, and regimes. To this, one might add genocidal conflicts to denote conflicts in which the belligerents regularly commit genocidal acts (not in the sense of conflicts which are intrinsically genocidal). These distinctions allow us to identify whether an individual genocidal act (such as an ethnic massacre) is an isolated incident or part of a cluster of such incidents that form a campaign. In turn, we can also determine whether a campaign of genocidal character is conducted in a limited fashion (perhaps in pursuit of a political or military objective), is an intrinsic part of the political project of a genocidal regime, or is part of a pattern of genocidal recidivism either by a single government or successive regimes in a country. This framework immediately allows us to distinguish between different kinds of ending and, significantly, how each implies different policy options.


How do we know what we know about the potential for international forces to end a threat of genocide or mass atrocities? The dominant model for drawing conclusions follows two paths. First, it refers to a very small number of cases in which action was taken and makes general conclusions about what happened and how this might apply elsewhere. Second, it turns to a broader selection of historical cases and shows what was not done. These negative examples are used to demonstrate how “we,” defined as those who might intervene, failed to take adequately robust action, at least not in good time.[5] In this analysis, the dark history of past cases provides a foil for our redemptive future actions. But history offers much more than a mirror to view of our failings or triumphs; the evidentiary record of actual cases demonstrates a broad range of forces—local, national, and regional—that contribute to ending atrocities.

Violence halted by the perpetrators once goals are met

The USSR is the definitive example of violence with an instrumental logic, in which both the beginning and ending of violent campaigns followed political decisions internal to the perpetrator regime. Josef Stalin was able to start and stop his terror campaigns and his deportations by decree. The Indonesian government’s mass killing of members of the Communist party (1965-1966) as well as the Chinese Cultural Revolution (1966-1976) both follow a similar logic of use of mass violence against civilians in order to suppress ideological and political resistance. A comparable case is the Ethiopian Red Terror (1977-78). The revolutionary government of Mengistu Haile Mariam halted this campaign when it had successfully repressed the urban insurrection that had prompted the mass killings.

In the case of the Nigerian civil war, in which the federal government sought successfully to repress the secessionist region of Biafra, overtly genocidal statements by some generals threatened a post-war genocide. Ultimately such actions did not happen when the army won the war: as soon as the federal army achieved military victory, the killing stopped. Under the banner of ‘no victor, no vanquished,’ the Nigerian government immediately began a process of normalization and reconciliation. This was aided by the fact that, during the war, there had been no consensus within the government on the war aims—some militants were set on punishing the secessionists beyond their military defeat, while others were intent solely on winning the war and restoring Nigeria’s territorial integrity. The latter group won out and the violations that had accompanied the war (and indeed provoked the separatist bid) rapidly ended.[6]

However, for the victims/ survivors and their descendants, the injustice remains very much alive as they seek recognition, compensation, and reparation. In many cases, the remnant indigenous people retain a tenacious hold of their experience as victims of genocide, which is the basis for them making claims against contemporary governments.

Colonial and settler genocides are important instances in which the ‘ending’ is defined primarily by the fact that the dominant power—the colonialists—got what they wanted. These cases are varied, particularly in the degree to which the indigenous peoples were able to survive, physically and socio-culturally. Centuries-long processes of expropriation, as well as the removal and destruction of groups through killing, hunger, disease and demoralization, may be said to have ended when the target groups were compelled to submit (or, in the limiting case of Tasmania, when they were eradicated). However, for the victims/survivors, and their descendants, the injustice remains very much alive, as they seek recognition, compensation and reparation. In many cases, the remnant indigenous people retain a tenacious hold of their experience as victims of genocide, which is the basis for them making claims against contemporary governments.

The case of Guatemala is a variant of this, a combination of a settler genocide and a counterinsurgency genocide. The war in Guatemala (1960-1996) was among the bloodiest of Latin America’s Cold War conflicts. Two years in the 1980s stand out as the most lethal phase of the conflict. Between 1981-1983, some 100,000-150,000 Guatemalan Maya were killed by the national armed forces.[7] As part of a scorched earth counter-insurgency plan that made use of long-standing racist assumptions about Maya and capitalized on their geographic and social isolation, government forces killed, raped, tortured, and forcibly displaced Maya in the rural mountain regions. Beginning in 1983, the army undertook measures to control the survivors, ushering in a second phase of assault marked by a combination of amnesty and intensified militarization of surviving communities. In the worst hit community, Rabinal, 14.6% of the population was killed and 99.8% of the victims were Maya.[8] Ending this intensive phase of the longer war, occurred only when the military leaders deigned that they had, as Suzanne Jonas quoted a military leader, “killed them enough,” meaning that the Maya communities were brought sufficiently under control.

Elite dissension or exhaustion within the perpetrator regime

Elite dissension or exhaustion recurs as a factor in many of the cases presented. The 1992 military campaign in the Nuba Mountains in Sudan is one such example.[9]  The ruling National Islamic Front was divided on its war aims, with some extremists (mostly party cadres) determined on the complete socio-cultural transformation of the Nuba people, and a pragmatic group (led by army officers) determined solely on containing or defeating the Sudan People’s Liberation Army (SPLA) rebellion. A combination of the internal politics of the government and the ruling party, the exhaustion of the military effort, and growing discontent among the wider northern Sudanese population about the excesses perpetrated during the jihad led the pragmatic group to win the day. The result was that, when the genocidal campaign of 1992 was ended, it was replaced not by another campaign of comparable ambition, but a nasty low-intensity counterinsurgency. The war continued, associated with ongoing violations, but no further genocidal campaigns. However, Khartoum evidently retained the capability to mount further campaigns of extreme violence, and the logic of the Sudanese political crisis continued to indicate the possibility that it might respond to a new political-military challenge in precisely such a manner.

In the case of East Pakistan/Bangladesh during the war of 1970-1971, divisions among Pakistani government leaders contributed to a considerable lessening of the violence even before the Indian invasion decisively ended the killings. The implication is that the genocidal campaign in East Pakistan was already waning, though it required the invasion to decisively bring an end to the violence targeted at the East Bengali elite. The Indian military victory brought about the independence of Bangladesh and the definitive termination of the prospects of further genocide. However, the new state was itself born in turmoil and did not achieve stability for many years.

Siad Barre’s campaign in northwest Somalia that reached a climax in 1988 is another case in which the exhaustion of the government’s military capacity combined with elite dissension to bring a genocidal campaign to an end. In this case, the dissension took the form of mutinies by military units and militia, unhappy at their own exclusion from power. What began as a counterinsurgency against the Somali National Movement rebels and their sympathizers, and escalated into genocidal onslaught against the Isaaq clan family, turned into the disintegration of both government and rebellion and the replacement of institutionalized armed forces with fragmented clan-based militia. The genocidal campaign ended in anarchy, and the state collapse that followed bred further genocidal campaigns by some of the militia groups that then seized power at a local level. (The campaign against the Bantu peoples of the Jubba Valley, introduced below, is one example, sadly not the only one.)

Victims of violence flee or otherwise resist

An important set of endings occurs through flight and asylum—when the target population gets itself out of harm’s way, or is evacuated. Catherine Besteman’s work on the Somali Bantus is a prime example.[10] During the Somali civil war after 1989, the Bantu populations of the Jubba Valley, long a marginalized and powerless group, were targeted by well-armed militia associated with different military clan factions. Their land was seized, their villages destroyed, and many Bantu men were killed and women raped. Virtually all took refuge in neighboring Kenya. Some tried to return to Somalia but faced ongoing violence. The ultimate solution for many Bantus has been resettlement in the United States. In a sense, the mass asylum of the Bantus is an admission of defeat—they are unlikely ever to return to their homeland. But it is a realistic and humane course of action in the face of desperately limited options.

In some cases, armed resistance by groups identified with the target group has ended genocide. One case is Yoweri Museveni’s National Resistance Army (NRA) in Uganda which ended the massacres in the Luwero Triangle in 1983-1984. The Ugandan NRA succeeded in fighting President Milton Obote’s Ugandan National Liberation Army (UNLA) to a standstill (aided by the death of Obote’s able chief of staff in a helicopter accident). But the overthrow of that regime and the collapse of its short-lived successor, leading to Museveni’s military victory and accession to power, did not bring an end to violence in Uganda. The groups most associated with the perpetration of the Luwero massacres (the Acholi and Langi soldiers of the UNLA) were the focus of a long-running insurgency in northern Uganda and provided the leadership of the Lord’s Resistance Army, which has mounted a campaign of fear across four countries.

The ending of a genocidal regime in Rwanda in July 1994 has not ended genocidal conflict across the Great Lakes region in the subsequent fourteen years, but actually amplified it.

Another instance is the military campaign of the Rwandese Patriotic Front (RPF) that ended the Hutu Power’s genocidal regime in Rwanda in 1994. While the RPF victory did undoubtedly bring the genocide to a decisive end, this case is also not unproblematic. Arguably, the RPF invasion in 1990 contributed to the radicalization of elements of the Hutu leadership who endorsed a campaign of mass killing. Moreover, the great majority of the massacres occurred in the first three weeks of the genocide, so that by the time of the RPF victory and the French Operation Turquoise, the level of killings had passed their peak. Most important, however, is the observation that although Rwanda has not witnessed genocide since the RPF victory, the consequences of the genocide and the means whereby it was ended include the war in the D.R. Congo, which has included genocidal violence on a large scale. The ending of a genocidal regime in Rwanda in July 1994 has not entailed the ending of genocidal conflict across the Great Lakes region in the subsequent fourteen years, but actually amplified it.

In the case of the Nuba Mountains of Sudan, resistance by the SPLA against enormous military odds—was sufficient to halt a genocidal campaign, which at its peak during the jihad of 1992 threatened the complete socio-cultural eradication of the Nuba people as a distinct group. But this resistance was not sufficient to overthrow the regime or weaken it to the point of exhaustion. Since 1992 there have been at least two episodes in Sudan that arguably qualify as genocidal campaigns (the clearances of the Upper Nile oilfields in the late 1990s and the Darfur counterinsurgency of 2003-2004), and most recently a return to war in the Nuba Mountains itself (in June 2011). Thus, although the Nuba SPLA may have defeated a genocidal campaign, they did not bring to an end conflict involving massive violence against civilians in Sudan.

Interventions waged by interested outside parties

Military intervention has ended genocide and mass atrocities in some cases: for example, Allied powers against Nazi Germany (1945), India against Pakistan in Bangladesh (1971), Vietnamese forces against the Khmer Rouge in Cambodia (1978), and the Tanzanian-led defeat of Idi Amin in Uganda (1979). Saddam Hussein’s massively violent campaigns against the Iraqi Kurds and Marsh Arabs were decisively ended by the U.S.-led invasion of 2003. It is important to note that, in none of these cases, was ending genocide either the main motivation or the immediate objective of the military action, which was undertaken when the regimes in question attacked or destabilized their neighbors or threatened the interests of a great power. In each of these cases, the interventions came very late in the day. In Bangladesh, most of the targeted killing had ceased some months before the Indian military intervention that was prompted by the exodus of Bangladeshi refugees into India in the context of long-standing Indian-Pakistani political rivalry and on-off armed conflict. In the case of Cambodia, the Vietnamese invasion was prompted by Cambodian military aggression against Vietnam, and humanitarian aims were entirely secondary to the Vietnamese military action. Although Saddam Hussein’s human rights record was cited as justification for the U.S. invasion, the immediate reasons were associated with fear of his possession of weapons of mass destruction in the context of the U.S. “war on terror.”

In none of these cases did armed intervention bring about immediate stability. There may (or may not) have been an improvement, but no matter how deplorable the regime that was overthrown, the invasion that followed was not a panacea. Germany, along with Eastern Europe, after 1945 was divided; surviving Jews in some previously occupied areas suffered post-war pogroms; and the largest modern population transfers occurred after the war when some 12 million ethnic Germans were moved back into Germany. This is not the equivalent to genocide, but does bear testimony to the instability following the military defeat of Nazi Germany. Bangladesh had a turbulent first few years as an independent nation. Considerable violence followed the Vietnamese victory over the Khmer Rouge and Cambodia did not enjoy peace for many years. Many Ugandans consider the conflict and genocide that followed the defeat of Idi Amin to have been worse than his dictatorship. And following the toppling of Saddam Hussein and subsequent American occupation of Iraq, there was extensive violence against civilians and ethnic cleansing.

Humanitarian interventions?

The war in Bosnia-Herzegovina is often cited as a case in which international armed intervention produced a decisive ending. Although a substantial UN peacekeeping force (often misidentified as a civilian protection force) with a mandate that extended to protecting humanitarian operations was deployed to Bosnia almost from the beginning of the conflict in 1992, it was not until 1995 that international forces truly confronted the Bosnian Serb army. This was only after the key objectives of the Bosnian Serbs—which entailed genocidal campaigns throughout areas they had designated for the Serb Republic—had already been accomplished. But the ending cannot simply be attributed to the NATO bombing campaign. Other key factors include political fracture between the Bosnian Serbs and the Serbian political leadership under Slobodan Milosevic (despite ongoing military aid from the Serbian army) and a counter-offensive by Bosnian and Croatian forces, reorganized and rearmed with international support, that fundamentally changed the military geography and decisively showed that Serb territorial consolidation had reached its limit. The Dayton Agreement negotiated in 1995 effectively ended the war, but in reaching beyond an immediate end to the violence, it established a system of governance precisely based on the political-ethnic definitions and divisions of the conflict. The Dayton formula has thus far proved incapable of re-uniting Bosnia-Hercegovina in any meaningful sense and has therefore cemented rather than remedied the ethnic cleansing of the war years.

The Dayton Agreement negotiated in 1995 effectively ended the war, but in reaching beyond an immediate end to the violence, it established a system of governance precisely based on the political ethnic definitions and divisions of the conflict.

The case of Kosovo in 1999 is an important anomaly and precedent, in which NATO’s military action was intended to avert an impending genocide.[11] The circumstances of this case remain highly controversial: while it is clear that Serbian government-aligned forces carried out systematic atrocities and forceful deportation of Kosovo’s ethnic Albanian population, some researchers have questioned the role of the Kosovo Liberation Army, alleging that it provoked an escalation of violence in order to bring about an international military intervention. It is also evident that ethnic cleansing did not cease with the NATO military occupation, the difference being only that on this occasion it was the Kosovars who forcibly expelled the Serbs.

The above is but a sampling of cases one might examine to better understand the local, national and regional forces, the political contexts, and the range of circumstances possible in the ending of mass atrocities and genocide.


On June 17, 2009, General Scott Gration, U.S. Presidential Envoy for Sudan, stated that Darfur was experiencing “remnants of genocide,” and thereby touched off a bitter disagreement within the Obama Administration, notably with U.S. Ambassador to the United Nations Susan Rice. Two days earlier, Rice had described the situation as “genocide,” and earlier that month President Obama had used the words “ongoing genocide”.[12] Journalists’ accounts of the disagreement used the adjective “furious” to describe Rice’s response to Gration’s comments.

Empirically, there is no question that military campaigns that might qualify as potentially “genocidal” had ceased in early 2005, possibly earlier.[13] By 2009, the violence showed a distinctly different pattern of multiple sources of threat, and mortality rates in the refugee and displaced persons camps were largely back to normal levels.  Yet there remained an enormous, vulnerable population of displaced civilians whose way of life had been irretrievably changed, and who remained insecure and fearful. The government retained the capacity and will to conduct large-scale campaigns of violence against civilian groups. What, precisely, had “ended” and what was “ongoing”? How would the answer to these questions frame the policy options that leaders might entertain?

On October 19, 2009, the debates within the Obama Administration were patched up with the announcement of a new Sudan policy. They retained the Bush Administration’s use of “genocide” to describe the situation while taking no position on whether it had ended or was ongoing. The policy consisted of three simultaneously—and apparently equally weighted—priorities: a “definitive end to conflict, gross human rights abuses, and genocide in Darfur,” implementation of the 2005 Comprehensive Peace Agreement between Sudan’s North and South, and efforts to ensure that Sudan would not again become a haven for terrorists.

The debate between Gration and Rice was neither simply semantic nor was it purely a disagreement over policy options. It related to a question that neither policy experts nor researchers have adequately engaged: how does one recognize when genocide or mass atrocities end? This question, let alone the equally, if not more complicated, question of how such violence ends has been caught up in the normative assumption about how they ought to end (i.e. international armed interventions that rescue the innocent from certain annihilation).

The ethical imperative announced in this “ought” has produced at best-confused policies and at worst policies that aggravated the likelihood that atrocities would occur. When the question of response is translated into the application of an unassailable ethical imperative, too frequently differences in the political and historical contexts of mass atrocities are ignored. The lines are then drawn in the name of separating the good from the bad, obscuring what knowledge is needed to provide the basis for effective policy. Study of the full array of contexts, actors, and measures for engaging potential or actual mass atrocities are thereby foreclosed.

The ethical imperative announced in this “ought” has produced at best confused policies and at worst policies that aggravated the likelihood that atrocities would occur.

An optimal ending of mass atrocities—including the rescue of vulnerable civilians, the punishment of perpetrators, and a just reconstruction of the state to address the conditions that enabled violence to occur—rarely, if ever, is achieved. But perhaps even more dangerous than the failure to achieve such endings is the way its very conception blocks understanding of what actually has been (and might be) achieved. Actual endings are suboptimal. There is no celebration for any example where thousands (if not more) are brutally murdered. But when we pay attention to what forces have played an actual role in ending violence, we can better understand how to interact with complex situations, what instruments might achieve which specific goals, and what can realistically be expected in terms of ending mass atrocities when this goal is clouded by other agendas.

Some endings are the successful completion of a genocidal campaign, called off when the perpetrator regime is consolidated or when the political landscape alters such that the political rationale for mass violence against civilians is reduced. Other endings occur when the genocidal regime is removed from power, through successful resistance or invasion, or is fought to a standstill. Possibly, these endings may prefigure another configuration of mass violence, carried out by members of the former target group against their erstwhile oppressors. This was the case in Uganda and has continued to be a pattern throughout the Great Lakes.

Some endings are simply respite—a gap between genocidal campaigns conducted by a regime with an intact apparatus of mass violence. Such a regime may have an enduring genocidal ideology (and hence motive) or may have other political or military objectives that translate intermittently into genocidal intent and genocidal campaigns. In cases of respite—such as the settler-colonial genocides, Stalin’s Soviet Union and Saddam’s Iraq—a genocidal campaign can be re-started at any time, against the same or different target groups. The Sudanese case is a variant of the latter, in which the government responds violently to diverse threats, sometimes with mass violence against civilians and sometimes not.

Another important distinction is between asymmetric and symmetric conflicts. Where there is a clear dominant power, such as an imperial state that is involved in repressing relatively powerless minorities, the conflict is asymmetric. In these circumstances, the state may possess the capacity to start and stop genocidal campaigns. It is the clear victor and determines the outcome from start to finish. By the same token, a dramatic change within the ruling elite of that state can lead to a decisive end to genocides, as was the case with Stalin’s death. Where conflicts are more symmetric and there is no such dominant power, as in the ethnically-defined state and sub-state polities common in central Africa, the prospects are different. In such cases, the ending of one genocidal campaign may be the occasion for the targeted group to mobilize to mount a counter-genocide against its former oppressors.

The concept of a “genocidal regime” implies a state with sufficiently institutionalized machinery to be able to plan and implement genocidal campaigns in a centrally-directed manner. While this is axiomatically the case for Nazi Germany and also holds for many other cases under discussion (notably Rwanda, Guatemala and Saddam Hussein’s Iraq), it is not the case for the D.R. Congo and Somalia. In these two cases the state has either collapsed entirely or cannot exercise basic functions including fielding an army. It is also questionable whether the concept of a genocidal regime holds for a number of weak and fragile states including Uganda (especially during the 1980s), Burundi, and possibly (controversially) Sudan. In such cases, “genocidal conflict” is a useful term that designates a dysfunctional political system in which political groups (often ethnically defined) pursue their objectives using violence, which often becomes sufficiently extreme and group-targeted that it qualifies as genocide.

Current models for weak and fragile states identify these countries as deviations from a Weberian norm of institutionalized political authority—they are defined by what they are not. Such analysis fails to pay attention to what actually happens in such countries, and how some of them successfully avoid or resolve conflict, despite not building strong institutions. It is preferable to analyze such countries in terms of the political processes which actually lead to certain outcomes.

One framework, proposed by Alex de Waal in his analysis of Sudan and its neighbors to the west (Chad and Central African Republic) is that of a “political marketplace” that functions in accordance with socio-cultural rules and dispenses patrimony.[14] Under this framework, provincial elites (in command of local constituencies that are usually ethnically-defined) seek to maximize the price of their loyalty, while metropolitan elites (in government) seek to drive down the price of such loyalty. This political marketplace is managed using violence, in times of both peace and war—with the result that the distinction between peace and war becomes blurred, and “peace agreements” may not actually lead to an end to violence. Knowing the rules of the political game, the Sudanese elites are remarkably civil among one another, even while violence is the order of the day in the provinces. Occasionally, an insurgent or a governing elite attempts a “game changer” that is liable to unleash a round of violence, amplified by an order of magnitude. It is in such cases that genocidal violence is most likely to occur. De Waal identified six episodes in the Sudanese civil wars since 1983 in which such game-changers have been attempted. This implies that ending genocidal campaigns in Sudan is, initially, a matter of stabilizing the political marketplace, to be followed by an effort aimed at addressing the structural conflict in Sudan that causes both chronic violence and recurrent genocidal campaigns.

It is disturbing to note that in the Great Lakes, Balkans, and Trans-Caucasus, members of many ethnic groups articulate a version of history which emphasizes how they were historic victims of genocide, and how the inevitable response to this victimhood is to organize to inflict similar violence on the former perpetrators. These histories become self-justifying and self-fulfilling charters for genocidal violence. Interventions at any level in such cases need to be attentive to the layers of historical arguments and how they are deployed for political purposes.

The contemporary project of preventing genocide—a modest aim of stopping only the most heinous crime against humanity—has become fused with much more ambitious goals of decisively resolving conflict and achieving transitional justice. Because genocide is defined as uniquely evil, we tend to infer that the response should be the establishment of a new regime, with ethical credentials commensurate with the horror that preceded it. This is a historical naiveté. Our approach eschews arguing from ‘ought’ to ‘is’, and instead addresses the complexities of real politics and develops a rich comparative evidence base. Rather than deriving analysis and policy from universals drawn from moral impulses, it seeks practical responses in the details of particular circumstances. Such an approach promises not only to be truer to reality but also to provide the tools for more practically effective policies for prevention and reaction that should achieve better results.

This article appears in the print edition of the Fletcher Forum of World Affairs. Vol. 36, No. 1 (Winter 2012), pp. 15-31

About the Author

Alex de Waal is the Executive Director of the World Peace Foundation at the Fletcher School. Jens Meierhenrich is Senior Lecturer of International Relations at the London School of Economics and Political Science. Bridget Conley-Zilkic is Research Director at the World Peace Foundation at the Fletcher School.

About the Journal

Founded in 1975 and published biannually, The Fletcher Forum of World Affairs is the student-managed foreign policy journal at The Fletcher School of Law and Diplomacy. The publication provides a broad, interdisciplinary platform for analysis of legal, political, economic, environmental, and diplomatic issues in international affairs.


[1] “End of an Era in Libya: Qaddafi is Killed in Sirte,” Al Arabiya, October 20, 2011, (accessed December 12, 2011).

[2] Rania Al Gamal, “Clues to Gaddafi’s Death Concealed From Public View,” Reuters, October 22, 2011, (accessed December 12, 2011).

[3] David Scheffer, “The Future of Atrocity Law,” Suffolk Transnational Law Review 25 (2002): 389-432.

[4] See, for example, Alex Bellamy’s work comparing the atrocity prevention agenda with that of more traditional peacebuilding practices. Alex Bellamy, “Mass Atrocities and Armed Conflicts: Links, Distinctions, and Implications for the Responsibility to Prevention,” (Washington, D.C.: The Stanley Foundation, 2011), (accessed December 12, 2011)

[5] The most influential presentation of this approach can be found in Samantha Power, A Problem from Hell: America and the Age of Genocide (New York: Basic Books, 2002).

[6] Sam Amadi, “Colonial Legacy, Elite Dissension and the Making of Genocide: The Story of Biafra,” Social Science Research Council Webforum: How Genocides End, January 10, 2007, (accessed December 12, 2011).

[7] Suzanne Jonas, “Guatemala: Acts of Genocide and Scorched-Earth Counterinsurgency War,” in Samuel Totten and William Parsons, eds., Century of Genocide (New York and London: Routledge, 2009), 381.

[8] Etelle Higonnet, ed., Quiet Genocide: Guatemala: 1981-1983 (New Brunswick and London: Transaction Publishers, 2009), 27.

[9] Alex de Waal, “Averting Genocide in the Nuba Mountains.” Social Science Research Council Webforum: How Genocides End, December 22, 2006, (accessed December 12, 2011).

[10] Catherine Bestemann, “Genocide in Somalia’s Jubba Valley and Somali Bantu Refugees in the U.S.,” Social Science Research Council Webforum: How Genocides End, April 9, 2007, (accessed December 12, 2011).

[11] Then U.S. Ambassador-At-Large for War Crimes, David Scheffer, referred to “indicators of genocide” at a press briefing ten days into NATO’s Kosovo campaign. President Clinton used similar language on June 25, 1999. See Samantha Power (2002), 468.

[12] Kristina Wong, “US tries to walk back from comments downplaying genocide in Darfur,” broadcast on ABCNews, June 18, 2009, (accessed December 12, 2011).

[13] Alex de Waal, Chad Hazlett, Christian Davenport and Joshua Kennedy, “Evidence-Based Peacekeeping: Exploring the epidemiology of lethal violence in Darfur,” Harvard Humanitarian Initiative, Working Paper 3, March 2010,

[14] Alex de Waal, “Dollarized,” London Review of Books, 32 (12) (June 24, 2010): 38-41.